Comparative Reliability of 108 Scales and their Short-Form Counterparts

Short-form scales are often necessary for large omnibus surveys. This study compared the reliability of 108 short-form scales and single-item indicators included in the New Zealand Attitudes and Values Study (NZAVS) with their full-form parent counterparts. Scale psychometrics were evaluated using an omnibus dataset pooling multiple samples with a foundation sample that employed a planned missing design presenting a random selection of 66% of the 1,012-item pool to n=2,947 undergraduates. Comparative reliability estimates for all short and full-form scales are reported. The majority of short-form scales displayedadequate reliability, indexed by coefficient omega. Fit statistics for all 108 full-form scales are also reported, and factor loadings are provided in an online supplement. These results calibrate conclusions for studies that necessarily employ short-form scales, and indicate that the majority of short-form scales employed by the NZAVS display adequate internal reliability relative to their original full-form (parent) versions.